CA AB 2013
Generative Artificial Intelligence: Training Data Transparency
Requires GenAI developers to publish documentation about training datasets including sources, data types, copyright status, personal information inclusion, and processing methods.
Jurisdiction
California
US-CA
Enacted
Sep 28, 2024
Effective
Jan 1, 2026
Enforcement
Not specified (likely CA AG under unfair competition law)
Signed September 28, 2024; effective January 1, 2026
Who Must Comply
Exemptions
National Security/Military
high confidenceNational security/military/defense systems (federal entities only)
Conditions:
- • Federal entity
- • National security purpose
Aircraft Operation
high confidenceSystems for aircraft operation
Conditions:
- • Aircraft operation purpose
Security/Integrity Systems
high confidenceSystems solely for security or integrity purposes
Conditions:
- • Security/integrity purpose only
Safety Provisions
- • High-level summary of training datasets
- • Sources or owners of datasets
- • Alignment with intended purpose
- • Number and types of data points
- • Copyright/trademark/patent/public domain status
- • Whether datasets purchased or licensed
- • Personal information or aggregate consumer data inclusion
- • Synthetic data usage
- • Data collection timeframes
- • Cleaning/processing methods
Compliance Timeline
Jan 1, 2026
Training data documentation must be posted on website
Enforcement
Enforced by
Not specified (likely CA AG under unfair competition law)
Penalties
Not specified in statute
Quick Facts
- Binding
- Yes
- Mental Health Focus
- No
- Child Safety Focus
- No
- Algorithmic Scope
- Yes
Why It Matters
Addresses training data transparency concerns. Relevant for IP/copyright discussions. Relatively light requirements (documentation only).
Recent Developments
Signed September 2024. Focuses on transparency about training data - relevant to copyright and data governance debates.
Cite This
APA
California. (2024). Generative Artificial Intelligence: Training Data Transparency. Retrieved from https://nope.net/regs/us-ca-ab2013
BibTeX
@misc{us_ca_ab2013,
title = {Generative Artificial Intelligence: Training Data Transparency},
author = {California},
year = {2024},
url = {https://nope.net/regs/us-ca-ab2013}
} Related Regulations
CA SB 942
Requires large GenAI providers (1M+ monthly users) to provide free AI detection tools, embed latent disclosures (watermarks/metadata) in AI-generated content, and offer optional manifest (visible) disclosures to users.
CA SB 53
First US frontier AI transparency law. Requires large AI developers (>$500M revenue) to publish governance frameworks, submit quarterly risk reports, and report critical safety incidents. Applies to models trained with >10^26 FLOP.
NY RAISE Act
Requires large AI developers of frontier models operating in New York to create safety protocols, report critical incidents within 72 hours, conduct annual reviews, and undergo independent audits. Creates dedicated DFS office funded by developer fees.
VT AADC
Vermont design code structured to be more litigation-resistant: focuses on data processing harms rather than content-based restrictions. AG rulemaking authority begins July 2025.
NE AADC
Nebraska design code blending privacy-by-design with engagement constraints (feeds, notifications, time limits) aimed at reducing compulsive use.
AR HB 1958
Requires all Arkansas public entities to create AI policies with mandatory human-in-the-loop for final decisions. Covers state departments, schools, and political subdivisions.